Main

PPHN is a disorder of term and near term infants with significant morbidity and mortality. PPHN results from disruption of the normal decrease in PVR at birth. PGI2, the main cyclo-oxygenase product of arachidonic acid in the vascular tissue, is a potent vasodilator and its actions are mediated by cAMP (1,2). It is one of the important mediators of the decrease in PVR at birth (3,4). PGI2 has been used widely in the treatment of adults with PH. Although there are reports of PGI2 being used anecdotally in human infants (5,6), systematic studies on the dose response effects of PGI2 in PPHN are lacking. Milrinone selectively inhibits PDE3, resulting in accumulation of cAMP in myocardium and vascular smooth muscle, improving myocardial performance and producing vasodilation. Milrinone has been used in patients to improve pulmonary hemodynamics in association with systemic hemodynamic dysfunction (7,8). Studies on the use of milrinone in neonates are lacking. Figure 1 shows the PGI2–cAMP signal transduction pathway in the vascular smooth muscle. PGI2 and milrinone in combination can act synergistically in increasing cAMP levels and hence enhance the relaxation of the vascular smooth muscle.

Figure 1
figure 1

Signal transduction pathway of PGI2–cAMP system. PGI2 and milrinone increase cAMP by different mechanisms and hence combination of these two therapies will greatly increase cAMP levels enhancing the relaxation of the vascular smooth muscle. COX, cyclo-oxygenase; AC, adenylate cyclase.

Full size image

The newborn pulmonary circulation is different from the adult pulmonary circulation in its anatomy and physiology (9,10). The response of the infant with PPHN to NO is closer to that of newborn lambs with persistent pulmonary hypertension induced by ligation of the ductus arteriosus (11) than it is to the response of adult humans with PH. Examination of the responses to manipulating cAMP with PGI2 and milrinone in such lambs seemed warranted in advance of clinical trials. Thus we studied the dose response to IT PGI2, the response to intravenous milrinone and enhancement of the effect of PGI2 by milrinone in newborn lambs with PH induced by ligation of the ductus arteriosus.

METHODS

The study was performed at the Center for Developmental Lung Biology at the State University of New York (SUNY) at Buffalo. In a well-established lamb model of PPHN, ligation of the ductus arteriosus 9 d before delivery causes PPHN (12–14). These studies were approved by the Institutional Animal Care and Use Committee (IACUC) at SUNY, Buffalo.

Fetal ductal ligation was performed via thoracotomy at 129 d of gestation (term = 146 d) as previously described (13). Time dated pregnant ewes (n = 6) were brought to the laboratory animal facility 24–72 h before surgery. The ewe was intubated after intravenous pentothal (750 mg) and ventilated with 0.5–1% isoflurane in oxygen. A left lateral thoracotomy was performed in the fourth intercostal space of the fetus, and the ductus arteriosus ligated. The fetal was then returned to the uterus and the abdominal wall was closed. The ewe was allowed to recover for 9 d. At 138 d gestation, the ewe was anesthetized and the fetus was partially exteriorized via hysterotomy, the carotid artery and jugular vein were exposed, and polyvinyl catheters were inserted and advanced into the aorta and right atrium, respectively. A left thoracotomy was performed and polyvinyl catheters placed in the main pulmonary artery (PA) and left atrium (LA). An ultrasonic flow transducer (Transonics Systems Inc., Ithaca, NY) was placed around the main PA. The lamb was then intubated by direct visualization and the endotracheal tube secured. The lambs were delivered and placed under an infant warmer to maintain temperature at 39°C. Lambs were covered in a plastic bag to prevent heat loss. Ventilation was initiated with a time cycled pressure controlled Sechrist ventilator (Sechrist Industries Inc., Anaheim, CA) at a rate of 60, positive inspiratory pressure (PIP) of 24–26 cm of H20, positive end-expiratory pressure (PEEP) of 4–5 cm of H20, Inspiratory time (I-time) of 33%, and fraction of inspired oxygen (Fio2) of 1.0. Initial stabilization included IV fluids at 100 mL/kg/d, monitoring blood pressure, heart rate, temperature, oxygen saturation, arterial blood gases (ABG), glucose, and electrolytes. The lambs were sedated with fentanyl infusion (3 μg/kg/h) and also received fentanyl bolus of 5 μg/kg i.v. when needed. The lambs were paralyzed with pancuronium (0.1 mg/kg q 3–4 prn] at birth and prn. Lambs received normal saline or whole cord blood for low mean systemic blood pressure (SBP) of <35 mm Hg and bicarbonate for base deficit of ≥8 mEq/L. Whenever the lambs received supportive therapy such as blood products, bicarbonate and volume, we deferred initiation of PGI2 protocols for at least 30 min, to minimize the effects of these therapies on hemodynamic variables. ABG were done frequently during stabilization to maintain PaCO2 between 35 and 55 mm Hg and pH between 7.35 and 7.45. Fio2 was kept at 1.0 during the study. All catheters were connected to a physiologic recorder (Gould Instrument Systems, Inc., Valley View, OH) for monitoring of SBP, PAP and left atrial pressure (LAP). The flow probe was connected to the flow meter (Transonic Systems Inc., Ithaca, NY) for monitoring of pulmonary blood flow (PBF). PVR was calculated using the standard formula: PVR = (PAP – LAP)/PBF.

Delivery of prostacyclin and milrinone.

PGI2 (Sigma Chemical Co., St. Louis, MO) was prepared fresh immediately before the start of the experiment using glycine buffer. PGI2 was delivered as a bolus dose directly into the trachea. The doses of PGI2 studied were 200, 500, 1000, 2000, 4000, and 8000 ng/kg given in random order. Glycine buffer (vehicle) was used as a control. All IT doses were prepared to a final volume of 2.5 mL. The doses were delivered by the feeding tube, cut to exact length, so that the doses are delivered 0.5–1 cm beyond the tip of the endotracheal (ET) tube. Autopsy studies show that the ET tube tip is generally 4 cm above the carina. Hemodynamic changes were recorded before and after administration of each dose. A 30-min interval was maintained between doses to allow for the PAP to come back to the baseline (pre-dose values). On completion of all of the doses of PGI2, milrinone (Baxter Healthcare Corporation, Deerfield, IL) was administered as a bolus dose of 100 μg/kg i.v. over 5 min followed by an infusion at 1 μg/kg/min. Baseline hemodynamic measurements were recorded before and for an hour after administration of milrinone. PGI2 was then repeated randomly by IT route at the same doses while on intravenous milrinone infusion. Hemodynamic measurements and ABG were done before and after the IT bolus dose. These measurements were made continuously to determine the onset and duration of action of PGI2. Onset of action was defined as the time for any decrease in PAP after administration of PGI2. Duration of action was defined as the time elapsed since administration of PGI2 for the PAP to decrease and then come back to its original baseline. Experiments were done in six newborn lambs with PPHN. Data are expressed as mean ± SEM, with n representing the number of animals. Statistical comparisons of the curves were performed with one way or repeated measures ANOVA as appropriate. Fisher's post hoc test was used as needed to compare among groups. All statistical analysis was performed with StatView software (Abacus Concepts, Berkley, CA). Significance was accepted at p < 0.05.

RESULTS

Tables 1 and 2 summarize the data of PVR, PAP, PBF, and SBP after randomly administered IT PGI2 and intravenous milrinone when given separately (Table 1) and in combination (Table 2). The baseline PAP before administration of PGI2 doses was between 60 and 65 mm Hg (Table 1). Percent changes in pulmonary hemodynamics are presented in Figure 2. PGI2 significantly decreased PAP at all doses but the vehicle did not (Fig. 2A). There was a significant increase in PBF with increasing doses of IT PGI2 (Fig. 2B). IT PGI2 decreases PVR from predose baseline values in a dose dependent manner (Fig. 2C). Vehicle did not affect hemodynamics and PGI2 did not affect SBP.

Table 1 Hemodynamic variables following randomly administered IT PGI 2 or intravenous milrinone in newborn lambs with pulmonary hypertension induced by antenatal ductal ligation
Full size table
Table 2 Hemodynamic variables following combination of IT PGI 2 and intravenous milrinone in newborn lambs with pulmonary hypertension induced by antenatal ductal ligation
Full size table
Figure 2
figure 2

Dose response effects of randomly administered IT PGI2 on percentage change in mean PAP (A), PBF (B), and PVR (C) in newborn lambs with fetal ductal ligation induced pulmonary hypertension. Data represent mean ± SEM of six newborn lambs. *p < 0.05, ANOVA repeated measures; **p < 0.01; †p < 0.001 compared with predose values, Fisher's Post hoc test.

Full size image

Intravenous milrinone by itself decreased PVR by 22% (Fig. 3). Intravenous milrinone increased pulmonary blood flow by 18% (Fig. 3). The effect of intravenous milrinone on mean PAP and mean SBP were not significant. IT PGI2 decreased mean PAP and PVR in the presence of intravenous milrinone in a dose-dependent manner (Fig. 4, A and C). Intravenous milrinone enhanced the response to PGI2. At doses of 500 and 1000 ng/kg PGI2 reduced mean PAP to a greater extent in the presence of intravenous milrinone than it did alone (Fig. 5A). Similarly, PGI2 produced a greater decrease in PVR in the presence of intravenous milrinone at doses of 200 ng/kg and 500 ng/kg than it did alone (Fig. 5B). Milrinone facilitates the response to PGI2 at lower doses.

Figure 3
figure 3

Percentage change in PVR, mean PAP, mean systemic blood pressure, and pulmonary blood flow after intravenous milrinone (100 μg/kg bolus followed by infusion of 1 μg/kg/min for 60 min) in newborn lambs with fetal ductal ligation induced pulmonary hypertension. Data represent mean ± SEM in six lambs. *p < 0.05 compared with predose values, Fisher's post hoc test.

Full size image
Figure 4
figure 4

Percent change in PAP (A), pulmonary blood flow (B), and PVR (C) of randomly administered IT PGI2 in the presence of intravenous milrinone (bolus = 100 μg/kg; infusion = 1 μg/kg/min) in newborn lambs with fetal ductal ligation induced pulmonary hypertension. Data represent mean ± SEM of six newborn lambs. *p < 0.05, ANOVA repeated measures; ‡p < 0.05, **p < 0.01; †p < 0.001 compared with predose values, Fisher's post hoc test.

Full size image
Figure 5
figure 5

Comparison of percent change in PAP (A) and PVR (B) of IT PGI2 alone (dashed line) with IT PGI2 plus intravenous milrinone (solid line) in newborn lambs with fetal ductal ligation–induced PH. Data represent mean ± SEM of six newborn lambs. IT PGI2 doses of 500 and 1000 ng/kg significantly reduced PAP in the presence of intravenous milrinone compared with IT PGI2 alone. Similarly IT PGI2 doses of 200 and 500 ng/kg reduced PVR in the presence of milrinone. *p < 0.05 compared with similar dose of IT PGI2; Fisher's post hoc test.

Full size image

The onset of action of IT PGI2 on PAP was seen within 60 s of administration of the drug. The range of onset of action varied from 39–60 s (49.7 ± 10.1 s) for PGI2 alone compared with 10–20 s (15.4 ± 4.1 s) for PGI2 in the presence of intravenous milrinone (Fig. 6A). The onset of action of PGI2 was significantly shorter in the presence of milrinone. The duration of action of PGI2 was significantly prolonged in the presence of intravenous milrinone compared with PGI2 alone (Fig. 6B). The range of duration of action was 5–7 min (6.13 ± 1.01 min) for PGI2 alone compared with 8–11 min (9.36 ± 1.18 min) for PGI2 in the presence of intravenous milrinone.

Figure 6
figure 6

Onset (A) and duration (B) of action of PGI2 alone (open squares) and in the presence of intravenous milrinone (shaded triangles). The onset of action was significantly shorter and the duration of action was significantly longer in the presence of milrinone. †p < 0.05 by ANOVA repeated measures vs PGI2 alone. *p < 0.05 and **p < 0.01 by Fisher's post hoc test vs corresponding PGI2 doses.

Full size image

No significant difference in mean SBP was noted before and after IT PGI2 at all doses studied. Similarly, there was no difference in mean SBP before and after IT PGI2 in the presence of milrinone. Lower doses of PGI2 produced only minimal change (0% to 2%) in mean SBP compared with a 6–12% decrease at higher doses (doses >2000 ng/kg) both in the absence and presence of intravenous milrinone. The data on mean PAP to mean SBP ratio is shown in Figure 7. PAP to SBP ratio was significantly lower following IT PGI2 compared with predose (Fig. 7A). Similarly, IT PGI2 in the presence of milrinone had a significantly lower PAP/SBP ratio compared with predose (Fig. 7B). This effect was more pronounced at lower doses (<2000 ng/kg of PGI2). IT PGI2 alone or with intravenous milrinone did not produce any significant difference in arterial blood gas parameters including Pao2, pH, and Paco2 before and after various doses of PGI2.

Figure 7
figure 7

Changes in PAP/SBP ratio following IT PGI2 with and without intravenous milrinone. (A) Represents PAP/SBP ratio before and after IT PGI2. (B) Represents PAP/SBP ratio before and after IT PGI2 in the presence of intravenous milrinone. Mean values are given and error bars represent SE. □ (open squares) represents the ratio before PGI2 and ▴ represents data post IT bolus of PGI2. *p < 0.05 by Fisher's post hoc test; †p < 0.05 by ANOVA repeated measures.

Full size image

DISCUSSION

Closure of ductus arteriosus of the fetal lamb 7–10 d before delivery causes persistent PH after birth (13). In our experiments, mean PAP soon after birth before start of the study were between 60–65 mm Hg. Ductal ligation also induces an increase in the proportion of partially and fully muscularized pulmonary arteries at the level of the terminal bronchiole and within the acinus (14). Thus, the anatomic and physiologic changes in the ductal ligation model of PPHN are similar to alterations reported in human neonates dying with idiopathic PPHN (14). This model of PH of the newborn has been used extensively in the preclinical studies of inhaled nitric oxide (iNO) on pulmonary hemodynamics and survival (12,15,16).

The effects of aerosolized PGI2 in PH in adults are well studied. Aerosolized PGI2 has been shown to be at least as effective as inhaled nitric oxide in decreasing PH in both animals and humans (17,18). Aerosolized PGI2 selectively dilates the pulmonary circulation and redistributes PBF away from nonventilated lung regions (19) and has been shown to improve oxygenation in patients with ARDS or acute lung injury (19–21). Aerosolized PGI2 reduced pulmonary pressures and improved right ventricular stroke volume in patients with PH undergoing cardiac surgery (22). The data on the effects in the neonate are anecdotal (5,23,24). IT instillation of PGI2 at 50 ng/kg when injected as a bolus dose in four preterm infants with documented PH resulted in a significantly improved oxygenation index without systemic hypotension (23). Neither human nor animal studies have addressed the role of PGI2 as a first line therapy in the management of PPHN soon after birth nor has there been determination of the pulmonary hemodynamic response to various doses of PGI2 in neonates.

Most of the studies in adults with PH have delivered PGI2 by intravenous or aerosolized route. As noted above, PGI2 was given by IT instillation in a pilot study of neonates with apparent success. Each of these methods of drug delivery to the lung has drawbacks in distribution, deposition with aerosols or systemic delivery with intravenous administration. As we were doing the first dose response studies in this very labor-intensive model of fetal ductal ligation induced PPHN, we chose the IT route to be certain of the dose deposited in the lung. Despite potential limitation on distribution, PAP responses were seen within a minute in all of our studies and lasted 5–11 min (PGI2 half life of 90–120 s) of administration of the drug, indicating the effectiveness of the IT route of administration in these experiments. The doses of PGI2 used were extrapolated from clinical and animal studies of inhaled PGI2. In our study, all doses of IT PGI2 effectively and selectively reduced PAP and PVR with minimal effect on SBP. As PAP and resistance decreased and pulmonary blood flow increased, it is probable that systemic blood flow and oxygen delivery increased. We failed to demonstrate any improvement in oxygenation with PGI2 either alone or with milrinone. This could be related not only to the significant intracardiac shunting of blood from right to left via the foramen ovale secondary to significant PH but also to intrapulmonary shunting in an instrumented, sick postoperative newborn lamb. Also, IT administration of the drug preferentially deposits the drug to a poorly ventilated, dependent part of the lung, limiting its effects on oxygenation.

In adult patients following cardiac surgery milrinone increased cardiac index and reduced pulmonary artery occlusion pressure and systemic vascular resistance (25). It decreased pulmonary arterial and venous vascular tone without increasing cardiac work or impairing pulmonary oxygenation in hypoxic dogs (26). A single 50 μg/kg i.v. bolus of milrinone produced a 31% reduction in PVR, a 42% increase in cardiac output, and a 12% reduction in mean PAP in adult patients with severe left ventricular dysfunction (27). Inhaled milrinone did not affect systemic arterial pressure in cardiac surgical patients with pulmonary hypertension (28). Milrinone as a loading dose of 50 μg/kg followed by an infusion of 0.5 μg/kg/min decreased PVR and SVR and increased cardiac index in neonates with low cardiac output following cardiac surgery (29). Dipyridamole, a PDE5 inhibitor had significant hemodynamic effects in both the pulmonary and systemic circulations of newborn lambs with PH (12). In our experiments, intravenous milrinone by itself decreased PVR and increased PBF with no significant affect on SBP. Our data demonstrate a good response in a lamb model of PPHN.

The use of PDE inhibitors for maintenance of the PGI2 induced second messenger cAMP may offer the possibility of prolonging and increasing the vasodilatory effect of nebulized PGI2 in the lung vasculature. A graphic depiction of decrease in PAP following IT PGI2 (2000 ng/kg) from a representative study both in the presence and absence of intravenous milrinone is shown in Figure 8. In our studies, the onset of action was significantly shorter and the duration of action was significantly longer with PGI2 in the presence of intravenous milrinone compared with PGI2 alone. Co-administering PGI2 with a systemic PDE3 inhibitor not only produced a greater decrease in PVR at lower doses, but also prolonged the action of PGI2 on PAP. The shorter time to onset of action (measured in seconds) may not be clinically relevant but its longer duration of action (measured in minutes) may be of clinical interest. The short half-life of PGI2 makes it imperative to administer it as a continuous inhalation for sustained clinical benefit. Our study validates the use of this combination therapy in a newborn lamb model of PH. Administration of PGI2 by IT bolus is clinically impractical, but its administration by continuous nebulization along with intravenous milrinone seems relevant and practical in certain clinical situations.

Figure 8
figure 8

Graphic depiction of a representative study demonstrating the effects of PGI2 on PAP. The kinetic information about the duration of response to PGI2 alone (A) is compared with same dose of PGI2 in the presence of intravenous milrinone (B). Intravenous milrinone shortens the onset and prolongs the duration of action of PGI2. The numbers adjacent to the recording indicate mean PAP in mm Hg.

Full size image

Several studies have addressed the role of systemic PDE inhibitors in combination with inhaled PGI2 in adults (30–33). Doses in our experiments are similar to the doses used in the clinical management of patients with cardiac failure and cardiogenic or septic shock. Subthreshold systemic doses of monoselective PDE3 (motapizone), PDE4 (rolipram) and PDE5 (zaprinast), and dual-selective PDE 3/4 inhibitors cause significant amplification of the pulmonary vasodilatory response to inhaled PGI2, while limiting the hypotensive effect to the systemic circulation (31). Co-administration of PDE inhibitors with inhaled iloprost, a prostacyclin analogue, markedly enhanced the prostanoid-induced pulmonary artery pressure decrease while maintaining the lung selectivity of the vasodilatory response (32). The absence of data on cardiac output makes it difficult to comment on vascular selectivity in our experiments. The lack of cardiac output and systemic vascular resistance (SVR) measurements is a significant limitation to the study, which is a result of a conscious decision to not perform double thoracotomies (to measure cardiac output) in these sick newborn lambs. In the absence of SVR data we used PAP/SBP ratio as a marker of vascular selectivity. Even though a declining PAP/SBP ratio is helpful, it does not prove that SVR did not drop significantly. PAP to SBP ratio decreased with PGI2 both in the presence and absence of intravenous milrinone. The ratio was lower at lower doses when PGI2 was administered alone or with milrinone. However, the ratio increases at higher doses of PGI2 with a concomitant decrease in SBP and this effect could be exaggerated by intravenous milrinone. This observation has clinical implications. The clinician should avoid higher doses of PGI2 especially when it is co-administered with intravenous PDE3 inhibitor.

There is paucity of data on PGI2 with PDE3 inhibitors in newborn human or animals. We have demonstrated that IT PGI2 and intravenous milrinone each decrease PVR in a well-established newborn lamb model of PPHN. Systemic administration of milrinone in combination with IT PGI2 produced a synergistic response in reduction in PVR, more so at clinically relevant doses of 200 and 500 ng/kg (extrapolates to 20–50 ng/kg/min over 10 min by aerosol). Combination of systemically administered milrinone with lower doses of PGI2 may enhance the pulmonary hemodynamic response with probably minor and clinically not relevant side effects on the systemic vasculature. These findings may be of relevance in the clinical management of infants with pulmonary hypertension.